NEWMEN
← back to all models

Qwen: Qwen3 30B A3B Instruct 2507

qwen/qwen3-30b-a3b-instruct-2507

Alibaba (Qwen)open-weight131K context6 providersIntelligence 58.0· est.

Cheapest provider

$0.05 / 1M

StreamLake

Fastest provider (p95)

188 tok/s

Nebius

Intelligence (estimated)

58.0

Family + generation + popularity + price tier

Per-provider performance

Latency / throughput / uptime / price measured across providers over the last 30 minutes of live traffic. This is what proves “sourced cheapest” — Atlas mode draws on these per call to serve the cheapest path that holds quality.

ProviderQuantInput $/1MOutput $/1MLatency p50 / p95Throughput p50 / p95Uptime 30mSuccess
StreamLakeundisclosed$0.0482$0.19312308ms / 4877ms55 / 75 tok/s100.00%100.0%
SiliconFlowq8· fp8$0.0900$0.30001667ms / 7673ms42 / 91.2 tok/s100.00%99.1%
AtlasCloudq8· fp8$0.1000$0.30002562ms / 5861ms8 / 44 tok/s99.20%99.3%
Nebiusq8· fp8$0.1000$0.3000353ms / 974ms101 / 188 tok/s99.86%99.2%
WandBfull· bf16$0.1000$0.3000307ms / 1268ms84 / 115 tok/s100.00%100.0%
Alibabaundisclosed$0.1300$0.5200638ms / 9106ms85 / 181.4 tok/s100.0%

“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).

Intelligence estimate

No public benchmark numbers indexed for this model yet, so the leaderboard score is derived from the catalogue data we sync: model family standing, generation, and live usage rank — applied identically to open- and closed-weight models. The heuristic ceiling sits below confirmed-benchmark frontier models so curated rankings stay clearly on top.

Want a curated score? File a benchmark report and we’ll add it to the next sync.

How Atlas mode sources Qwen: Qwen3 30B A3B Instruct 2507

  • Strict mode — pin Qwen: Qwen3 30B A3B Instruct 2507 exactly and we pass it straight through, sourced from the cheapest provider above. The same model, no substitutions — currently StreamLake at $0.05/1M.
  • Atlas mode — the default. Each call is auto-optimized for the cheapest path that holds quality, at least 5% off going direct from call one and climbing as it ramps. You always see which model served the call and exactly what you saved — thumbs-down anything you don’t like for a full refund.